Connection of a synchronous compensator to a power grid

Abstract

To connect a synchronous compensator (2) to a power grid (100), a connection condition for the output voltage is defined, identified by a connection frequency (1'B) and by a connection phase shift (9B) between the vector of the output voltage and the grid voltage. A starting condition is also defined, with a starting frequency (fy) and a starting phase shift (9A). The starting condition is calculated, with formulas as a function of the connection condition, in such a way that the output voltage, starting from the starting condition, reaches the connection condition by means of a pre-established freewheel slowdown transient of the synchronous compensator (2). After these calculations, the synchronous compensator (2) is started by means of a static converter (3). When the starting frequency and the phase shift (fA, 9A) are reached, the static converter (3) is disconnected from the synchronous compensator (2), which slows down, deterministically leading to the connection condition. Then, the synchronous compensator (2) is connected to the power grid (100).

Description

[0001] Title: "Connection of a synchronous compensator to a power grid".

[0002] DESCRIPTION

[0003] Technical field

[0004] The present invention is developed in the field of reactive power regulation in a power grid. In particular, the invention relates to a method for connecting a synchronous compensator to a power grid, and a synchronous compensator system.

[0005] State of the art

[0006] Synchronous compensators for regulating reactive power in a power grid are known in the art. A synchronous compensator is a synchronous electric machine, with direct current excitation, whose rotor, at least in steady state conditions, is left free to rotate without being connected to significant mechanical loads, nor to significant prime movers.

[0007] Therefore, when connected to the power grid, the synchronous compensator does not exchange significant amounts of active power, except to compensate for friction of the rotor and any flywheel. In fact, these rotate at a constant speed, synchronously with the grid frequency or with its submultiples, depending on the number of poles of the stator and rotor windings.

[0008] In a known manner, by adjusting the excitation current it is possible for the compensator to absorb reactive power from the grid, acting as an inductor, or generate reactive power, as a capacitor. In particular, the reactive power exchanged is proportional to the difference between the grid voltage and the voltage induced by the rotor magnetic field in the stator windings.

[0009] At steady state, the induced voltage and phase shift levels between the induced voltage and the grid voltage are maintained within predetermined operating intervals of the synchronous compensator, to avoid excessive currents that would cause mechanical and thermal stress.

[0010] Problems of the background art

[0011] When a synchronous compensator must be connected to the grid, the output voltage at the terminals of its stator windings should be brought as close as possible to the grid voltage in terms of frequency, rms (root mean square) voltage and phase. Incorrect values of even one of these parameters would lead the synchronous compensator to exchange with the grid higher currents than those expected in the machine's operating intervals in a very short time, resulting in mechanical and thermal stress such as to shorten the life of the machine.

[0012] Typically, it is therefore prescribed that the connection must occur in the following intervals:

[0013] - difference between rms output voltage and rms grid voltage less than 10%,

[0014] - slip, i.e., frequency difference between grid voltage and output voltage, less than 0.5% of grid frequency,

[0015] - phase shift between output voltage and grid voltage less than 5°.

[0016] Commonly, before connection to the grid, the synchronous compensator is started by means of a static frequency converter. In detail, the static converter can set the rotation speed of the rotor and the frequency of the output voltage.

[0017] However, a static converter sized to bring the synchronous compensator to grid voltage would have very high costs. Therefore, it is not usually possible to bring the synchronous compensator to satisfy all connection conditions directly by means of the static converter, for an immediate connection to the grid. Instead, before connecting the synchronous compensator to the grid, it must be disconnected from the static converter to preserve the static converter from the grid voltage.

[0018] During this temporary detachment, the synchronous compensator begins a freewheel slowdown, reducing the speed with respect to that set by the static converter. Therefore, the synchronous compensator must initially be brought to a speed greater than the grid frequency. Once the static converter has been disconnected, the correct output voltage value is set, by means of appropriate excitation, and the output frequency is monitored, which progressively decreases. When the output frequency is sufficiently close to the grid frequency, such that the set slip condition is met, the electrical connection to the grid is controlled and carried out by a parallel device, as long as the phase shift is also within the limit established by the reference parameter.

[0019] However, since the speed imposed by the static converter is chosen empirically, and there are no other particular conditions imposed at the beginning of the sequence, the phase with which the slowdown transient begins is determined by the random instant at which the detachment procedure from the static converter begins, and the phase shift that occurs when the output frequency reaches the grid frequency is also random.

[0020] As a result, there is no guarantee that the phase shift will transit in the correct interval before the compensator frequency exits the interval set by the slip parameter set in the parallel device.

[0021] In fact, in the prior art the connection occurs by trial and error. The synchronous compensator is brought to a speed higher than the grid frequency, and it is left to freewheel slowdown, hoping that the connection conditions are met by chance. Alternatively, the procedure is repeated. A parameter known as the synchronization success rate expresses precisely the probability that the connection conditions are met.

[0022] In the event of very low success rates, for example when no flywheel is provided, and therefore the rotor slows down quickly, it is common to increase the allowed slip parameter for the connection, in order to increase the probability that the phase shift will reach the correct interval in good time. However, this subjects the synchronous compensator to unwanted stresses (see IEC 60034-1, 2022).

[0023] JP 2022043417 discloses an example of a static synchronous converter with a control to avoid the mentioned problems.

[0024] Summary of the invention

[0025] The object of the present invention is to control the process of connecting a synchronous compensator to the power grid, so as to increase the synchronization success rate to values substantially equal to 1, i.e., make the connection process substantially deterministic, avoiding having to repeat multiple attempts in the synchronization process.

[0026] It is further an object of the invention to minimize the stress on the synchronous compensator caused by overcurrents during synchronization.

[0027] These and other objects are achieved, alternatively with respect to JP 2022043417, by a method for connecting a synchronous compensator to a power grid, and by a synchronous compensator system, according to any of the appended claims.

[0028] According to the invention, before the synchronization of the synchronous compensator, a starting condition is identified, to be imposed at the initial instant of the freewheel descent sequence, given by the optimal value of the voltage frequency and phase at the ends of the synchronous compensator, with respect to the power grid conditions, for example by comparison with the grid voltage vector. The starting condition is calculated as a function of the connection condition that the output voltage must reach, with a frequency corresponding to the grid frequency, preferably measured in terms of average value at the time of the synchronization request, and zero phase shift with respect to the grid voltage. The calculations take into account the specific dynamics of the synchronous compensator, which describes the law of freewheel descent, so that the synchronous compensator, starting from the starting condition, reaches the slip and phase intervals allowed by the parameters of the parallel device, by means of a pre-established slowdown transient.

[0029] Preferably, in order to calculate the starting condition, a preliminary characterization step of the synchronous compensator is envisaged, to determine one or more dynamic parameters thereof, including deceleration parameters.

[0030] Following this calculation, the synchronous compensator is brought to the starting condition by means of a static converter, and is disconnected therefrom to perform the slowdown transient. The transient therefore leads the synchronous compensator to reach the connection condition in a substantially deterministic manner, and when such a condition is reached, the connection to the power grid is performed, with frequency, phase shift and rms voltage substantially at the centre of the allowable tolerance intervals, and thus with the minimum stress for the synchronous compensator.

[0031] Advantageously, the number of attempts required for synchronization is reduced, i.e., the success rate is maximized, and the life of the synchronous compensator is extended.

[0032] Further features and advantages of the invention will be recognisable by a person skilled in the art from the following detailed disclosure of exemplary embodiments of the invention.

[0033] Brief disclosure of the figures

[0034] For a better understanding of the following detailed disclosure, some embodiments of the invention are illustrated in the accompanying drawings, wherein:

[0035] - Figure 1 schematically shows a static synchronous compensator according to a preferred embodiment of the invention,

[0036] - Figure 2 shows a graph on a time shape of an output frequency of the synchronous compensator,

[0037] - Figure 3 shows a graph on a time shape of an output voltage of the synchronous compensator,

[0038] - Figure 4 shows a graph on time shapes of the output frequency, for different choices of starting conditions according to the invention when a parameter k varies, as well as a time shape for a choice of a selected starting condition without applying the invention, and

[0039] - Figures 5 and 6 show a graph on a time shape of an excitation voltage of the synchronous compensator during two control steps.

[0040] DETAILED DESCRIPTION

[0041] A synchronous compensator system according to an aspect of the present invention is indicated overall with the number 1. Synchronous compensator system 1 herein is meant as the assembly of a synchronous compensator 2 and some additional components functional to its synchronization with a power grid 100, and in general its operation.

[0042] The power grid 100 is subject to an alternating grid voltage, preferably three- phase, with a rms grid value, a predetermined grid frequency fn, e.g., 50 or 60 Hz, and a grid phase variable over time.

[0043] Synchronization is meant as a series of steps of a method for connecting the synchronous compensator 2 to the power grid 100, according to another aspect of the invention, which bring the synchronous compensator 2 to operate in synchronism with the power grid 100, i.e., with the same frequency.

[0044] Some known features of a synchronous compensator 2, useful for clarifying some aspects of the invention, are summarized below.

[0045] The synchronous compensator 2, in a known manner, is identified by a rotating electric machine, and in particular a synchronous machine. In particular, the synchronous compensator 2 comprises a stator 21 and a rotor 22, coaxial with each other. The rotor 22 is rotatably connected to the stator 21, for example by means of bearings (not illustrated), to rotate about a rotation axis. Preferably, the rotor 22 is housed in a cavity of the stator 21.

[0046] The synchronous compensator 2 further comprises conductive excitation windings, configured to be fed with an excitation current Iein direct current. In a known manner, the excitation windings are mounted to the rotor 22, and can be arranged so as to identify a predetermined number of pairs of magnetic poles, when traversed by the excitation current Ie.

[0047] The synchronous compensator 2 also comprises stator windings with output terminals 23 connectable to the power grid 100. The stator windings are mounted to the stator 21 and are arranged to identify a predetermined number of pairs of magnetic poles, preferably equal to the number of pairs of poles of the excitation windings.

[0048] The synchronous compensator 2 is configured to generate an output voltage to the output terminals 23, preferably three-phase, with an output frequency f. For this purpose, in a known example, the stator windings can be connected to each other in a star, wherein each winding is connected between a respective terminal 23 and a common star centre. In another known example, the stator windings are connected in a delta, each between a distinct pair of terminals 23. Still in a known manner, feeding an excitation current Iewhile the rotor 22 is rotating generates an induced voltage in the stator windings, proportional to the excitation current Ieand rotation speed. Under no-load conditions, i.e., when the output terminals 23 are not connected to the grid 100 or to other external devices, the output voltage corresponds to the induced voltage. When the output terminals 23 are instead connected to the power grid 100 or to another device that imposes a voltage that does not correspond to the induced voltage, the stator windings deliver or absorb a predominantly reactive current.

[0049] It should be noted that the rotor 22 is not connected to mechanical loads, and therefore the stator windings do not continuously absorb active power, except to a limited extent to compensate for the rotation friction of the rotor 22. Furthermore, the rotor 22 is not connected to a prime mover, and thus the stator windings do not continuously deliver active power. Instead, the rotor 22 may optionally be fixed to a flywheel (not illustrated), to increase its rotational inertia.

[0050] It should also be noted that, when excitation current is present, the mechanical rotation frequency of the rotor 22 is directly proportional to the frequency of the induced voltage in the stator windings. Therefore, except for phenomena of loss of pitch between the output voltage and the induced voltage, the output voltage also has an output frequency f which is proportional to the rotational frequency of the rotor 22. The proportionality ratio between these frequencies, in a known manner, is given by the number of polar pairs of the synchronous compensator 2. Therefore, the rotation frequency of the rotor 22 is equal to the frequency f of the output voltage or a submultiple thereof according to a fixed ratio.

[0051] The synchronous compensator system 1 comprises a static converter 3, configured to start the synchronous compensator 2 by setting its rotor 22 in rotation. In particular, the static converter 3 is configured for the connection to the output terminals 23 and for generating a converter voltage, with an adjustable converter rms value and converter frequency. It should be noted that the converter rms voltage value cannot exceed a limit value, generally lower than the rms grid voltage value.

[0052] The static converter 3 can be identified for example by an inverter. In the preferred embodiment, the static converter 3 comprises a plurality of thyristors. Additionally or alternatively, the static converter 3 may comprise transistors.

[0053] The synchronous compensator system 1 also comprises a control system 4, comprising one or more control units which may be interdependent or substantially independent of each other in carrying out the functions described below. Reference will be made herein to a preferred embodiment of the control system 4, noting that different architectures of the control system 4 can be contemplated.

[0054] Preferably, the control system 4 includes a converter controller 41, which may be integrated in the static converter 3. The control system 4, and in particular the converter controller 41, is configured to control the operation of the static converter 3, by sending opening and / or closing signals to its thyristors and / or transistors.

[0055] The synchronous compensator system 1 then comprises an exciter 5 of known type, for example a static exciter or a rotary exciter, connected to the rotor windings and controllable by the control system 4 to feed a variable value of excitation current Ie. In particular, the preferred control system 4 comprises an exciter controller 42 configured to control the exciter 5, which may be integrated in the exciter 5.

[0056] In more detail, the exciter controller 42 is configured to receive excitation commands, and control the exciter 5 as a function of the received excitation commands.

[0057] Preferably, during synchronization, the excitation commands are generated and transmitted to the excitation controller 42 by the converter controller 41. In a known manner, after the completion of the synchronization, the excitation commands can be generated and transmitted to the excitation controller 42 by an external controller, not illustrated, for example a controller of a distributed control system (DCS).

[0058] The synchronous compensator system 1 comprises a parallel device 6, configured for the automatic connection of the synchronous compensator 2 to the grid 100 and to the static converter 3. In particular, the parallel device 6 comprises a grid connection element 61, for example a grid switch, configured to connect and disconnect the output terminals 23 to the power grid 100. Furthermore, the parallel device 6 comprises a converter connection element 63, for example a converter disconnector, configured to connect and disconnect the output terminals 23 to the converter 3.

[0059] The grid and converter connection elements 61, 62 are controllable by the control system 4, between connection and disconnection configurations. In particular, the preferred control system 4 comprises a parallel controller 43, which can be integrated in the parallel device 6.

[0060] Preferably, the parallel controller 43 is configured to enable and disable an automatic parallel mode, wherein the parallel controller 43 is configured to identify an optimal connection instant between the output terminals 23 and the grid 100, as specified below, and to perform such a connection by means of the grid connection element 61 at the identified optimal instant. Preferably, the parallel controller 43 is configured to receive an enable command, and to enable the automatic parallel mode upon receipt of the enable command. Preferably, the enable command is generated and transmitted to the parallel controller 43 by the converter controller 41.

[0061] Furthermore, the parallel controller 43 is preferably configured to receive manoeuvring commands, and on such occasion to control the parallel device 6 in such a way as to perform the manoeuvres which are the subject of the manoeuvring commands. The manoeuvring commands may include, for example, commands for disconnecting the output terminals from the grid 100, and commands for connecting and disconnecting the output terminals 23 from the converter 3. Preferably, during synchronization, at least some manoeuvring commands, such as connection and disconnection commands of the converter 3, are generated and transmitted to the parallel controller 43 by the converter controller 41. Instead, once the synchronization is over, it is preferred that at least some manoeuvring commands, such as disconnection commands from the grid 100, are generated and transmitted to the parallel controller 43 by the external controller.

[0062] Some steps of the method according to the invention concerning the operating working of the synchronization are now introduced. However, as clarified below, these are not the starting steps in order of time, as the synchronization operation is generally preceded by setting and / or characterization steps, described in more detail below.

[0063] In order to start the synchronous compensator 2, the start is a condition in which the synchronous compensator 2 is connected to the static converter 3 and is disconnected from the power grid 100.

[0064] The exciter 5 is then controlled to feed a first continuous excitation Ieito the rotor windings. Furthermore, the static converter 3 is controlled to generate a converter voltage with increasing frequency. In a known manner, by appropriately setting a frequency increase rate, the rotor 22 starts to rotate synchronously with the increasing converter frequency, or with a predetermined sub-multiple thereof. Therefore, in this step the frequency of the induced voltage remains substantially equal to the converter frequency.

[0065] The static converter 3 is controlled to increase the converter frequency to bring the output frequency f to a predetermined frequency value, called the starting frequency fA, as it is the starting point for the main synchronization steps. In this step, the output frequency can be controlled in open loop, as set by the converter 3, or it can still be monitored by means of a frequency detector.

[0066] According to an aspect of the invention, the starting frequency fA is not chosen arbitrarily during start-up, but is previously stored in the control system 4 before starting the synchronous compensator 2, as disclosed in more detail below. The starting of the synchronous compensator 2 must therefore bring the output frequency f up to the starting frequency value fA that has been established in advance.

[0067] The starting frequency fA is greater than the grid frequency fn, but less than or equal to a maximum frequency of the converter 3, e.g., 52 Hz or 62 Hz.

[0068] In addition to the starting frequency fA, according to an aspect of the invention, a starting phase shift 9A between the output voltage and the grid voltage is also defined and stored in the control system 4. The starting frequency fA and the starting phase shift 9A are collectively referred to as a starting condition.

[0069] In particular, once the starting frequency fA has been reached, the control system 4, and specifically the converter controller 41, controls the static converter 3 to keep the rotation speed of the rotor 22 constant. This step is not visible in the graphs of Figures 2-4 only because it has a very short duration with respect to the time scale considered.

[0070] The control system 4 comprises a first phase comparator, preferably integrated in the converter controller 41. The first phase comparator is configured to detect the phase shift 9 between the output voltage and the grid voltage. In particular, the first phase comparator is configured to detect said phase shift 9 when the output frequency is equal to the starting frequency fA. To this end, preferably the first phase comparator comprises voltage detectors connected to the grid 199 and to one between the converter 3 and the output terminals 23.

[0071] Therefore, in this step the control system 4, and in particular the converter controller 41, monitors the phase shift between the output voltage and the grid voltage, until the starting phase shift 9A is reached.

[0072] In the preferred embodiment, each phase shift disclosed herein is defined and / or detected as a phase shift between vectors. In particular, the grid voltage and the output voltage of the synchronous compensator 2 are alternating quantities that are representable in a known manner by vectors with complex numerical values. Such vectors can be generally computed starting from a sampling of the signal of interest.

[0073] The phase shift 9 is thus identified as the phase difference between the grid voltage vector and the output voltage vector, that is, the angle identified between the two vectors, or again the phase of a vector given by the difference therebetween, which is also a complex number, variable over time, with a real part and an imaginary part generally not zero, except instantaneously.

[0074] As disclosed in more detail below, upon reaching the starting phase shift, the synchronization procedure comprises some operations including disconnecting the synchronous compensator 2 from the static converter 3. This results in a slowdown transient of the rotor 22 of the synchronous compensator 2.

[0075] In addition to the starting condition, a connection condition for the output voltage is also defined and stored in the control system 4. It is of interest for the invention that the connection condition includes a connection frequency I'B of the output voltage and a connection phase shift 9B between the output voltage and the grid voltage. Generally, the connection condition also includes a rms connection value, which however is of less interest as it is a parameter that is already easily obtainable according to the prior art. It should be noted that the connection frequency fu may also be defined as a connection slip. Slip is meant herein as the difference in frequency between the output voltage and the grid voltage. Similarly, the starting frequency f can also be defined as a starting slip.

[0076] Preferably, the connection frequency fs is equal to the grid frequency fn, i.e., the connection slip is zero. Furthermore, the connection phase shift 9B is a zero phase shift, and the rms connection value equals the rm s grid voltage value.

[0077] For at least one portion of the slowdown transient, the control system 4 is configured to monitor the output voltage, and in particular the frequency f of the output voltage and its phase shift 9 with respect to the grid voltage. In the preferred embodiment, the parallel controller 43 comprises a second phase comparator for this purpose.

[0078] Similarly to the first phase comparator, the second phase comparator is configured to detect the phase shift 9 between the output voltage and the grid voltage. However, the second phase comparator is configured to detect such a phase shift during at least one portion of the slowdown transient, in particular when automatic parallel mode is enabled in preparation for the connection.

[0079] This second phase comparator thus operates under different slip conditions from the first phase comparator, and is subject to different technical requirements. In fact, the first phase comparator operates at maximum slip conditions, although in the preferred embodiment the maximum slip is not greater than 2 Hz, while the second phase comparator operates at minimum slip conditions, almost zero. However, embodiments in which the phase shift 9 is always detected by a single phase comparator are not excluded.

[0080] The second phase comparator preferably includes voltage detectors connected to the grid 100 and to the output terminals 23. The parallel controller 43 is preferably also configured to detect the grid frequency fn, the output frequency f, and the rms grid and output voltage values. These quantities can be detected by means of specific transducers, or can always be obtained starting from a sampling of the signals of interest, performed by the same voltage detectors which are already used to determine the phase shift.

[0081] When the output voltage reaches the connection condition, or differs therefrom by less than a pre-established set of tolerances, the control system 4, and in particular the parallel controller 43, is configured to command the grid connection element 61 to connect the output terminals 23 to the power grid 100. Thereby, the slowdown transient is interrupted and concluded, and the rotor 22 starts to rotate with substantially constant frequency, corresponding to the grid frequency fnor a predetermined sub-multiple thereof. For example, the set of tolerances can be that defined in the introductory section of this disclosure.

[0082] During the deceleration transient, the rotor 22 is preferably freewheeling. Therefore, the rotation frequency of the rotor 22, and thus the frequency of the induced voltage, decrease according to laws of a uniformly decelerated rotary motion, where the deceleration is determined by the rotation friction and by any other energy losses of the system, for example connected to induced currents.

[0083] According to an aspect of the invention, the starting condition is calculated in a setting step preceding the operating steps of the synchronization. In particular, the starting condition is calculated by means of a set of pre-established formulas as a function of the connection condition, such that the slowdown transient of the synchronous compensator 2 brings the output voltage, starting from the initial condition, up to reaching the connection condition. It should be noted that, although the frequency f of the output voltage slows down with respect to the starting frequency fA, and thus sooner or later necessarily reaches the connection frequency, an incorrect choice of frequency fA and / or starting phase shift 9A does not guarantee that the connection phase shift 9B is reached precisely in the time interval in which the output frequency f is in the interval around the value of fa allowed by the slip parameter set in the parallel device 6.

[0084] Equations of uniformly decelerated rotary motion, and their reformulations, such as inverse formulas, which describe the slowdown transient, are shown below. By imposing that these formulas result in reaching the connection condition, they allow to calculate the starting condition, or even plural permissible distinct starting conditions, before starting the synchronization.

[0085] In the following, the symbol co=27tf will be used to indicate the pulsation of the output voltage, and con=27tfnthe grid pulsation. Furthermore, the symbol 9 indicates the angular phase shift between the output voltage and the grid voltage, preferably between the output voltage and grid voltage vectors. In this disclosure, the term pulsation is used to indicate a quantity measurable in rad / s, while frequency is used to indicate quantities in Hz, but in the field, pulsations are also commonly called frequencies, once it has been clarified that rad / s is intended to be used as a unit of measurement.

[0086] Starting from the starting condition, identified by point A of the graphs of Figures 2-4, where there is a starting output pulsation coA=27tfA and a starting phase shift 9A, and assuming that for a certain time interval an angular deceleration a (assumed positive) is constant, the equations of the uniformly decelerated rotary motion can be written as follows. (co - con) At + aAt22t1) A= co + aAt

[0087] It should be noted that the system (1) can also be expressed in other equivalent forms, thus changing the representation of the system, but not its solution. For example, with appropriate modifications, the initial pulsation COA can be chosen to be used in the first equation, instead of the pulsation co of a generic instant, or it can be chosen to express both equations as a function of the angular acceleration a’=-a, which will be negative during the transient, instead of the deceleration a assumed positive.

[0088] The Applicant has observed that the angular deceleration a does not assume a single constant value for the entire slowdown transient, but, in an embodiment wherein a preferred example of simplification is operated, the slowdown transient can be divided into two intervals, of duration Ati and At2 respectively, wherein the angular deceleration a assumes two distinct constant values ai and 012, as illustrated in Figure 2 and 4. The point at which the connection condition is reached is indicated as B, and the connecting point between the intervals Ati and At2 is indicated as C.

[0089] Therefore, in such an embodiment, the system of equations (1) is partially applied for each of the two intervals, obtaining the systems of equations (2) and (3). In other words, the equations of decelerated rotary motion are applied with distinct parameters, in particular distinct deceleration parameters, in the intervals Ati and At2.

[0090] Therein, coc and COB indicate the pulsation in points C and B of the graph of Figure 2, i.e., the point C at which the interval Ati ends and the point B at which the interval At2 ends, given by the instant at which the connection occurs and the slowdown transient ends. Similarly, 9c and 9B indicate the phase shift in points C and B.

[0091] The two distinct angular decelerations ai and 012 occur for the reasons described below, with reference to a sequence of steps which are carried out, in a known manner, during the slowdown transient.

[0092] The beginning of the slowdown transient, i.e., the point of the graph of Figure 2 indicated with A, where the interval Ati begins, is identified by a step of interrupting the first excitation current by the exciter 5, as commanded by the control system 4 to the exciter 5. In particular, this command is sent from the converter controller 41 to the exciter controller 42, and from the latter to the exciter 5. Thereby, the synchronism is lost between the voltage of the static converter 3 and the rotation speed of the rotor 22, which begins to slow down.

[0093] Next, the control system 4 commands to disconnect the static converter 3 from the synchronous compensator 2. Preferably, this occurs by commanding the switching off of the thyristors or output transistors of the converter 3, the disconnection of the converter 3, with the opening of a converter power switch (not depicted in the figure), and subsequently the opening of the converter connection element 62. This series of steps occurs in a short time at the beginning of the interval Ati. The commands for carrying out these steps are preferably generated by the converter controller 41 and sent out partly to the converter 3 and partly to the parallel controller 43, which in turn commands the converter connection element 62.

[0094] Then, the control system 4 commands the exciter 5 to feed the rotor windings with a second continuous excitation current IC2. In particular, the command for this step is transmitted from the converter controller 41 to the exciter controller 42, and from the latter to the exciter 5. The second excitation current Ie2 is greater than the first excitation current Iei, so that the induced voltage in the stator windings is compatible with the static converter 3, before the point A, and compatible with the grid voltage, within reaching the point B. In particular, the second excitation current Ie2 is set in a manner such that the induced voltage has a rms value substantially equal to the rms grid value. In an example embodiment, the first excitation current Ieiis substantially equal to 20% of the second excitation current Ie2. The consequent shape of the output voltage is shown in the graph of Figure 3. It should be noted that, after the point A, with the cancellation of the excitation current Ieand the disconnection of the converter 3, the output voltage decreases exponentially.

[0095] Feeding the second excitation current Ie2 causes an increase in deceleration, which goes from ai to 012, and so this occurs in point C of the graph.

[0096] Finally, following point C, the automatic connection mode of the parallel device 6 is enabled, so that it is ready for the automatic connection between the output terminals 23 and the grid 100, when the connection condition occurs. Such an enabling occurs during the interval At2. The phase comparator of the parallel device 6 thus allows the control system 4, and in particular the parallel controller 43, to monitor the phase shift. Upon reaching the connection condition, the connection to the grid 100 occurs by means of the grid connection element 61, which identifies the point B.

[0097] This sequence of steps determines the two different deceleration values ai to 012, which are parameters connected to the characteristics of each individual synchronous compensator 2, and which are known as measurable beforehand. In particular, preferably the method comprises, before the setting step wherein the starting condition is calculated, also a step of characterizing the synchronous compensator 2, which involves performing one or more tests to determine one or more dynamic parameters of the synchronous compensator 2, to be applied in the set of formulas used for the calculation of the starting condition. The determined dynamic parameters, in particular, comprise among them the deceleration a, preferably both decelerations ai, 012 applicable in the intervals Ati and At2.

[0098] Other known parameters in the equations (2) and (3) are those identified by the connection condition, i.e., the pulsation C0B=C0n, and the phase shift 9B=0, or equivalently 0B=2k7i, with positive integer k.

[0099] A preferred manner of obtaining the starting condition, applying the mentioned simplification, is disclosed below. In this preferred manner, the starting phase shift is also chosen to be zero, 9A=0. This is preferable for the operation of the phase comparator, which monitors having reached the starting phase shift and then identifies the instant when the control system 4 will command the start of the slowdown transient. Therefore, in the whole of the slowdown transient, between the start of the transient and reaching the connection condition, the phase shift between the output voltage and the grid voltage completes a whole number of complete revolutions, without further portions of revolutions remaining, except for the permitted tolerances. Consequently, 0B-0A=2k7L.

[0100] Furthermore, the value of Ati is assumed to be known in the preferred solution. In fact, Ati has a duration which cannot fall below the minimum time necessary to sequentially carry out the aforementioned steps of zeroing the first excitation current lei, disconnecting the static converter 3, and feeding the second excitation current Ie2. In the preferred real synchronous compensators, the minimum duration of Ati is between 10 s and 30 s, in particular between 15 s and 23 s, preferably about 18 s. In the preferred solution Ati is thus selected equal to its minimum permissible value. In more detail, Ati can also be determined as one of the dynamic parameters of the characterization test.

[0101] Therefore, according to the preferred solution, the two systems of equations (2) and (3) are rewritten as follows.

[0102] In the systems (4) and (5), in addition to the main unknown quantity COA, the duration of the interval At2, and the phase shift 9c and the pulsation coc at the time C when the second excitation current Ie2 is fed, are also variables. Therefore, by fixing the value of k attributed to 9B, the systems (4) and (5) together include four variables and four equations. The equations of the system (4) are both of first degree, remembering that Ati is fixed. The system (5) instead includes a first degree equation and a second degree equation.

[0103] One possible solution is to substitute coc, as expressed by the second equation of the system (5), in the first equation of the system (4), to then substitute 9c, from such an equation, in the first equation of the system (5). This equation, for sufficiently high values of k, is solvable to obtain At2. With At2 known, COA is derived from the second equation of the system (5).

[0104] Such a solution therefore allows to obtain, as k varies, a family of starting conditions, all with zero initial phase shift 9A, and with different starting pulsations COA. From the starting pulsation COA, the starting frequency can be derived as f = 2TTCOA. It should be noted that these calculations allow solutions to be found even in the case where a grid frequency other than the nominal value is measured, for example other than 59 Hz or 69 Hz.

[0105] An alternative solution of the systems (4) and (5) may involve, for calculation purposes, dividing the interval At2 into two theoretical intervals At’2 and At’ ’2, at a point D. The point D is chosen so that the phase shift 9 makes a whole number of revolutions between D and B, and less than one revolution between C and D.

[0106] It should be noted that the solution obtained from the systems (4) and (5) can be subject to corrections to take into account some real dynamics of the system that may deviate slightly from the hypotheses assumed. For example, if the rotational inertia of the synchronous compensator 2 is relatively low, in particular in the absence of flywheels, a corrective could be introduced to take account of the fact that the second excitation current Ie2 is not reached in point C instantaneously, and the rising transient of the excitation current could therefore have a non-negligible duration, during which the deceleration is intermediate between ai to 012.

[0107] In another embodiment, a less simplified approach can be adopted with respect to dividing the pre-established transient into two intervals, both with constant angular deceleration.

[0108] In this example, the angular deceleration is still constant in the interval At2, in particular from the moment when the second excitation current Ie2 is reached, and thus the deceleration 012. Therefore, at least in this interval it is still preferred to use uniformly decelerated motion formulas as identified by the formula (5).

[0109] Instead, before the interval At2, one or more intervals are generally identified, among which at least one is characterized by a non-uniform deceleration. In the embodiment of Figures 5 and 6, an interval Atic, immediately preceding At2, is an interval which goes from the point Cl, where the excitation current feed resumes, to the point C2, where the excitation current reaches a value equal to the second excitation current Ie2. In fact, the excitation current does not instantly reach the second excitation current IC2, but grows as a ramp and, depending on various construction parameters, can take a few seconds, for example five seconds.

[0110] In this interval, therefore, the rotary motion does not have constant deceleration, but constant angular jerk, where the angular jerk is the constant temporal derivative of the acceleration (or deceleration). It is therefore possible to use decelerated motion formulas with a constant jerk, which can be analogous to those of uniformly decelerated motion, where the constant deceleration is replaced by the mean deceleration.

[0111] Still in the embodiment of Figures 5 and 6, an interval Atiaimmediately following the point A is contemplated, where the deceleration is not constant, but a motion where exponentially decelerated rotary motion formulas are applicable, in particular decelerated according to an exponential damping, in accordance with the shape of the excitation voltage.

[0112] This motion with exponential damping, if continued for a sufficient time, identifies a transient which tends to stabilize as uniformly decelerated motion. If the time required for this stabilization is significantly less than Ati, an interval Atib can be identified, between the intervals Atiaand Atic, where the uniformly decelerated motion formulas can be applied, with deceleration ai. Otherwise, as can be seen in Figure 6, the intervals Atiaand Atic may be consecutive to each other.

[0113] In the example of Figures 5 and 6, therefore, several different types of formulas are considered for two or three intervals before At2. A person skilled in the art may adopt greater or lesser simplifications to increase or reduce the number of intervals to be considered before At2. In any case, similarly to what is proposed for the implementation example of formulas (4) and (5), it is possible to first characterize the system to determine the motion parameters of the different intervals, and then, knowing the durations of all the intervals preceding At2, to obtain the values of At2 and of the starting frequency f such that the phase shift 9 between the output voltage and the grid voltage makes a whole number of complete revolutions.

[0114] In the following, other less preferred embodiments are briefly discussed, which allow the systems (2) and (3) to be solved with different hypotheses, and therefore to reach the output condition deterministically, with different numerical results of the starting condition. In an embodiment, the angular deceleration a could be assumed substantially constant over the entire slowdown transient, and then directly use the system of equations (1) in a simplified and analogous manner to what is illustrated for the systems of equations (2) and (3). Such an embodiment is generally less accurate, and could result in the connection condition not being reached, unless it is taken into account with other correctives of the acceleration variation, or unless the construction characteristics of a particular real synchronous compensator make the decelerations in the two intervals sufficiently similar.

[0115] In some embodiments, non-zero starting phase shifts 9A could be allowed, which when reached would still be detectable by the phase comparator to command the start of the slowdown transient. The systems (2) and (3) would in any case become solvable by imposing other constraints, for example by fixing a priori the starting pulse COA or the duration of the interval At2.

[0116] In still other embodiments, it could be chosen to increase the duration of Ati beyond the preset minimum value, by delaying some operations by the control system 4. In this case, additional constraints may be introduced to alternatively solve the system.

[0117] Returning to the embodiment of formulas (4) and (5) (but by analogy this is also applicable to the embodiment of Figures 5 and 6), as anticipated, a family of starting conditions can be obtained as k varies. Therefore, the starting frequency fA may be chosen from a plurality of candidate frequency values. Figure 4 shows different output frequency shapes f for different choices of k.

[0118] It should be noted that, based on the calculation method set forth, all the candidate frequency values satisfy the property whereby, if adopted as starting frequency, the phase shift between the output voltage and the grid voltage during the slowdown transient completes a respective whole number of full revolutions, which is different for all the candidate values.

[0119] It should be underlined that the static converter 3 may not be able to realize, or at least not with sufficient accuracy, all the candidate frequency values. In particular, the converter 3 can be configured, when controlled in such a sense, to adjust the converter frequency to a plurality of step frequencies with discrete values.

[0120] Therefore, in choosing the starting frequency fA among the candidate values, those candidate frequency values which are distant from each of the step frequencies by more than a predetermined threshold are preferably discarded. Furthermore, candidate values that are higher than the maximum frequency of the converter 3 are preferably discarded, e.g., 52 Hz or 62 Hz.

[0121] It should be noted that, assuming the grid frequency fnconstant, the starting condition for a specific synchronous compensator 2 could be calculated once and for all with a single characterization and setting step, and then stored so as not to be modified or recalculated further, but only used in operation to synchronize such a synchronous compensator 2. This initial calculation can be carried out by the control system 4, or it can be carried out externally to then store the result.

[0122] In the preferred embodiments, the characterization of the synchronous compensator 2 is performed once, but then the setting with the calculation of the starting condition can be repeated by the control system 4 before each synchronization, to obtain optimal results each time for a value of the grid frequency fnjust detected. In more detail, the characterization step allows to determine dynamic parameters such as ai, 012 and Ati, but also the optimal value of k for the synchronous compensator 2. Subsequently, when the connection of the synchronous compensator 2 to the grid is required, a setting step is carried out, wherein the grid voltage is measured and the starting condition is calculated, based on the measured grid voltage, and on the values of k and the dynamic parameters determined during the characterization. Finally, after the setting, the steps of operation are carried out, wherein the converter 3, the exciter 5 and the parallel device 6 behave as disclosed.

Claims

CLAIMS1. Method for connecting a synchronous compensator (2) to a power grid (100) having an alternating grid voltage with a grid frequency (fn), wherein the synchronous compensator (2) comprises a rotating electric machine having:- a stator (21),- a rotor (22), rotatably connected to the stator (21),- excitation windings, mounted to the rotor (22), and- stator windings, mounted to the stator (21) and having output terminals (23) connectable to the power grid (100), the synchronous compensator (2) being configured to generate an output voltage to the output terminals (23), wherein the method comprises:- defining a connection condition for the output voltage, identified by a connection frequency (fn) of the output voltage, equal to the grid frequency (fn), and by a connection phase shift (9B) between the output voltage and the grid voltage,- defining a starting condition for the output voltage, the starting condition being identified by a starting frequency (fA) of the output voltage, greater than the connection frequency (fs), and by a starting phase shift (9A) between the output voltage and the grid voltage, characterized in that the starting condition is calculated, by means of a set of pre- established formulas as a function of the connection condition, in such a way that the output voltage, starting from the starting condition, reaches the connection condition by means of a pre-established slowdown transient of the synchronous compensator (2), and in that the method further comprises:- after calculating the starting condition, starting the synchronous compensator (2) bymeans of a static converter (3), setting the rotor (22) in rotation so as to bring the frequency (f) of the output voltage to the starting frequency (fA),- when the frequency (f) of the output voltage reaches the starting frequency (fA), keeping a constant rotation speed of the rotor (22) by means of the static converter (3), until the starting phase shift (9A) is reached,- when the starting phase shift (9A) is reached, starting a slowdown transient corresponding to the pre-established transient, in which the static converter (3) is disconnected from the synchronous compensator (2), and- terminating the slowdown transient by electrically connecting the output terminals (23) to the power grid (199), when the output voltage differs from the connection condition by less than a pre-established set of tolerances.

2. Method according to claim 1, wherein the pre-established transient is a freewheel slowdown transient.

3. Method according to claim 1 or 2, comprising, before defining the starting condition, performing a characterization test of the synchronous compensator (2) to determine one or more dynamic parameters of the synchronous compensator (2), wherein the one or more dynamic parameters are applied in said set of pre-established formulas.

4. Method according to any one of claims 1 to 3, wherein the set of pre-established formulas for calculating the starting condition comprises equations of uniformly decelerated rotary motion, or reformulations thereof, which describe the slowdown transient and result in the connection condition being reached.

5. Method according to any one of claims 1 to 4, wherein:- starting the synchronous compensator (2) comprises feeding to the rotor windings a first continuous excitation current (Iei),- starting the slowdown transient comprises, in order, interrupting the first excitation current (Iei), disconnecting the static converter (3) from the synchronous compensator (2), and feeding to the rotor windings a second continuous excitation current (Ie2), preferably greater than the first excitation current (Iei).

6. Method according to the combination of claims 4 and 5, wherein the equations of uniformly decelerated rotary motion, or reformulations thereof, are applied at least after reaching the second excitation current (Ie2).

7. Method according to claim 6, wherein the equations of uniformly decelerated rotary motion, or reformulations thereof, are applied before and after the start of feeding the second excitation current (Ie2), with distinct parameters, in particular distinct deceleration parameters (ai, 012).

8. Method according to claim 6, wherein the set of pre-established formulas comprises, for one or more intervals before reaching the second excitation current (Ie2), non- uniformly decelerated motion formulas, in particular exponentially decelerated motion formulas and / or constant-jerk decelerated motion formulas.

9. Method according to any one of claims 1 to 8, wherein the starting phase shift (9A) and the connection phase shift (9B) are zero phase shift (9) conditions between the output voltage and the grid voltage.

10. Method according to claim 9, comprising, in the step of keeping the rotation speed of the rotor (22) constant by means of the static converter (3), and during the slowdown transient, monitoring a phase shift (0) between the output voltage and the grid voltage by means of one or more phase comparators.

11. Method according to any one of claims 1 to 10, wherein the starting phase shift (9A) and the connection phase shift (9B) are identified as phase differences between complex vectors representative of the output voltage and of the grid voltage.

12. Method according to any one of claims 1 to 11, wherein the starting frequency is calculated in such a way that in the pre-established transient, between the start of the pre-established transient and the achievement of the connection condition, a phase shift (9) between the output voltage and the grid voltage performs a whole number of complete revolutions.

13. Method according to any one of claims 1 to 12, wherein:- the static converter (3) is configured to adjust a converter frequency to a plurality of step frequencies with discrete values,- calculating the starting condition comprises calculating a plurality of candidate frequency values, and selecting the starting frequency (fA) among the candidate frequency values, discarding candidate frequency values that are distant from each of the step frequencies by more than a predetermined threshold.

14. Method according to the combination of claims 10 and 13, wherein the candidatefrequency values are calculated in such a way that, by adopting distinct candidate frequency values as the starting frequency (fA), between the start of the pre-established transient and the achievement of the connection condition, the phase shift (9) between the output voltage and the grid voltage performs different whole numbers of complete revolutions.

15. Synchronous compensator system (1), comprising:- a synchronous compensator (2) comprises a rotating electric machine having:- a stator (21),- a rotor (22), rotatably connected to the stator (21),- excitation windings, mounted to the rotor (22), and- stator windings, mounted to the stator (21) and having output terminals (23) connectable to a power grid (100), the synchronous compensator (2) being configured to generate an output voltage to the output terminals (23),- a control system (4), configured to:- store a connection condition for the output voltage, identified by a connection frequency (fa) of the output voltage, equal to a grid frequency (fn) of a grid voltage of the power grid (100), and by a connection phase shift (9B) between the output voltage and the grid voltage,- calculate and store a starting condition for the output voltage, the starting condition being identified by a starting frequency (fA) of the output voltage, greater than the connection frequency (fs), and by a starting phase shift (9A) between the output voltage and the grid voltage, wherein the starting condition is calculated, by means of a set of pre-established formulas as a function of the connection condition, in such a way that the outputvoltage, starting from the starting condition, reaches the connection condition by means of a pre-established slowdown transient of the synchronous compensator (2),- a static converter (3), configured to start the synchronous compensator (2) setting its rotor (22) in rotation,- a parallel device (6), configured to connect and disconnect the synchronous compensator (2) to the grid (100) and to the static converter (3), the parallel device (6) comprising at least one phase comparator, configured to monitor a phase shift (9) between the output voltage and the grid voltage, wherein the control system (4) is further configured to:- after calculating the starting condition, command to the static converter (3) to start the synchronous compensator (2), setting the rotor (22) in rotation so as to bring the frequency (f) of the output voltage to the starting frequency (f / \),- when the frequency (f) of the output voltage reaches the starting frequency (fA), command to the static converter (3) to keep a rotation speed of the rotor (22) constant and to monitor a phase shift (9) between the output voltage and the grid voltage by means of the at least one phase comparator, until the starting phase shift is reached, and- upon reaching the starting phase shift (9A), command the starting of a slowdown transient of the synchronous compensator (2) corresponding to the pre-established transient, in which the static converter (3) is disconnected from the synchronous compensator (2) by the parallel device (6),- wherein the parallel device (6), after the start of the slowdown transient, is configured to command the electrical connection between the output terminals (23) and the power grid (199), terminating the slowdown transient, when the output voltage differs fromthe connection condition by less than a pre-established set of tolerances.

Images